fix voxelization issue with structures containing H

app.py

@@ -52,26 +52,35 @@ def update(inp, file, mode, custom_resids, clustering_threshold):
                 return "pdb code must be 4 letters or Uniprot code does not match", ""
     identifier = os.path.basename(filepath)
     if mode == "All residues":
-        print('using all residues')
+        print("using all residues")
         ids = get_all_protein_resids(filepath)
-    elif len(custom_resids)!=0:
-        print('using listed residues', custom_resids)
-        ids=get_all_resids_from_list(filepath,custom_resids.replace(","," "))
+    elif len(custom_resids) != 0:
+        print("using listed residues", custom_resids)
+        ids = get_all_resids_from_list(filepath, custom_resids.replace(",", " "))
     else:
-        print('using metalbinding')
+        print("using metalbinding")
         ids = get_all_metalbinding_resids(filepath)
     print(filepath)
     print(ids)
     try:
         voxels, prot_centers, prot_N, prots = processStructures(filepath, ids)
     except Exception as e:
-        return "Error", f"""<div class="text-center mt-4"> Something went wrong with the voxelization, reset custom residues and other input fiels and check error message <br> <br> <code>{e}</code></div>"""
+        print(e)
+        return (
+            "Error",
+            f"""<div class="text-center mt-4"> Something went wrong with the voxelization, reset custom residues and other input fiels and check error message <br> <br> <code>{e}</code></div>""",
+        )
     device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
     voxels.to(device)
-    
+
     model = Model()
     model.to(device)
-    model.load_state_dict(torch.load("weights/metal_0.5A_v3_d0.2_16Abox.pth", map_location=torch.device("cuda" if torch.cuda.is_available() else "cpu")))
+    model.load_state_dict(
+        torch.load(
+            "weights/metal_0.5A_v3_d0.2_16Abox.pth",
+            map_location=torch.device("cuda" if torch.cuda.is_available() else "cpu"),
+        )
+    )
     model.eval()
     with warnings.catch_warnings():
         warnings.filterwarnings("ignore")
@@ -107,7 +116,6 @@ def update(inp, file, mode, custom_resids, clustering_threshold):
     )
 
 
-
 def read_mol(molpath):
     with open(molpath, "r") as fp:
         lines = fp.readlines()
@@ -171,7 +179,7 @@ def molecule(pdb, probes, cube):
         </div>
         <div class="px-4">
         <label for="pdbmetal" class="relative inline-flex items-center mb-4 cursor-pointer ">
-            <input  id="pdbmetal" type="checkbox" class="sr-only peer" checked>
+            <input  id="pdbmetal" type="checkbox" class="sr-only peer">
             <div class="w-11 h-6 bg-gray-200 rounded-full peer peer-focus:ring-4 peer-focus:ring-blue-300 dark:peer-focus:ring-blue-800 dark:bg-gray-700 peer-checked:after:translate-x-full peer-checked:after:border-white after:absolute after:top-0.5 after:left-[2px] after:bg-white after:border-gray-300 after:border after:rounded-full after:h-5 after:w-5 after:transition-all dark:border-gray-600 peer-checked:bg-blue-600"></div>
             <span class="ml-3 text-sm font-medium text-gray-900 dark:text-gray-300">Show PDB metals</span>
           </label>
@@ -311,13 +319,11 @@ def molecule(pdb, probes, cube):
     allow-top-navigation-by-user-activation allow-downloads" allowfullscreen="" 
     allowpaymentrequest="" frameborder="0" srcdoc='{x}'></iframe>"""
 
+
 def set_examples(example):
-    n,code, resids = example
-    return [
-        n,
-        code, 
-        resids
-    ]
+    n, code, resids = example
+    return [n, code, resids]
+
 
 metal3d = gr.Blocks()
 
@@ -325,40 +331,66 @@ with metal3d:
     gr.Markdown("# Metal3D")
     with gr.Tabs():
         with gr.TabItem("Input"):
-            inp = gr.Textbox(            placeholder="PDB Code or Uniprot identifier or upload file below", label="Input molecule"
+            inp = gr.Textbox(
+                placeholder="PDB Code or Uniprot identifier or upload file below",
+                label="Input molecule",
             )
             file = gr.File(file_count="single", type="file")
-            
+
         with gr.TabItem("Settings"):
             with gr.Row():
                 mode = gr.Radio(
                     ["All metalbinding residues (ASP, CYS, GLU, HIS)", "All residues"],
                     label="Residues to use for prediction",
                 )
-                custom_resids = gr.Textbox(placeholder="Comma separated list of residues", label="Custom residues")
+                custom_resids = gr.Textbox(
+                    placeholder="Comma separated list of residues",
+                    label="Custom residues",
+                )
             with gr.Row():
-                clustering_threshold = gr.Slider(minimum=0.15,maximum=1, value=0.15,step=0.05, label="Clustering threshold")
-                distance_cutoff = gr.Slider(minimum=1,maximum=10, value=7,step=1, label="Clustering distance cutoff")
+                clustering_threshold = gr.Slider(
+                    minimum=0.15,
+                    maximum=1,
+                    value=0.15,
+                    step=0.05,
+                    label="Clustering threshold",
+                )
+                distance_cutoff = gr.Slider(
+                    minimum=1,
+                    maximum=10,
+                    value=7,
+                    step=1,
+                    label="Clustering distance cutoff",
+                )
         btn = gr.Button("Run")
-    n = gr.Textbox(label="Label",visible=False)
-    examples = gr.Dataset(components=[n,inp, custom_resids],
-    samples=[["HCA2", "2CBA", ""],
-                                     ["Nickel in GB1 dimer", "6F5N", ""],
-                                     ["Zebrafish palmitoyltransferase ZDHHC15B PDB", "6BMS", ""],
-                                     ["Human palmitoyltransferase ZDHHC23 AlphaFold", "Q8IYP9", "280,273,263,260,274,277,274,287"]],
-                           )
+    n = gr.Textbox(label="Label", visible=False)
+    examples = gr.Dataset(
+        components=[n, inp, custom_resids],
+        samples=[
+            ["HCA2", "2CBA", ""],
+            ["Nickel in GB1 dimer", "6F5N", ""],
+            ["Zebrafish palmitoyltransferase ZDHHC15B PDB", "6BMS", ""],
+            [
+                "Human palmitoyltransferase ZDHHC23 AlphaFold",
+                "Q8IYP9",
+                "280,273,263,260,274,277,274,287",
+            ],
+        ],
+    )
     examples.click(fn=set_examples, inputs=examples, outputs=examples.components)
-    #gr.Markdown(
+    # gr.Markdown(
     #    """ <small>Inference using CPU-only, can be quite slow for more than 20 residues. Use Colab notebook for GPU acceleration</small>
-    #"""
-    #)
-
+    # """
+    # )
 
     gr.Markdown("# Output")
-    
+
     out = gr.Textbox(label="status")
     mol = gr.HTML()
-    btn.click(fn=update, inputs=[inp, file, mode, custom_resids, clustering_threshold], outputs=[out, mol])
+    btn.click(
+        fn=update,
+        inputs=[inp, file, mode, custom_resids, clustering_threshold],
+        outputs=[out, mol],
+    )
 
 metal3d.launch(share=True)
-

utils/helpers.py

@@ -89,7 +89,7 @@ def get_all_protein_resids(pdb_file):
         prot = Molecule(pdb_file)
     except:
         exit("could not read file")
-    prot.filter("protein")
+    prot.filter("protein and not hydrogen")
     return prot.get("index", sel="name CA")
 
 
@@ -112,12 +112,13 @@ def get_all_metalbinding_resids(pdb_file):
         prot = Molecule(pdb_file)
     except:
         exit("could not read file")
-    prot.filter("protein")
+    prot.filter("protein and not hydrogen")
     return prot.get(
         "index",
         sel="name CA and resname HIS HID HIE HIP CYS CYX GLU GLH GLN ASP ASH ASN GLN MET",
     )
 
+
 def get_all_resids_from_list(pdb_file, resids):
     """Return all metal binding residues from a pdb file
 
@@ -131,7 +132,7 @@ def get_all_resids_from_list(pdb_file, resids):
     Returns
     -------
     resids : numpy.ndarray
-        indexes of name CA resids 
+        indexes of name CA resids
 
     """
 
@@ -139,12 +140,13 @@ def get_all_resids_from_list(pdb_file, resids):
         prot = Molecule(pdb_file)
     except:
         exit("could not read file")
-    prot.filter("protein")
+    prot.filter("protein and not hydrogen")
     return prot.get(
         "index",
         sel=f"name CA and resid {resids}",
     )
 
+
 def compute_average_p_fast(point, cutoff=1):
     """Using KDTree find the closest gridpoints
 

utils/voxelization.py

@@ -136,7 +136,9 @@ def voxelize_single_notcentered(env):
             voxelsize=0.5,
             validitychecks=False,
         )
-    except:
+    except Exception as e:
+        print(e)
+        print(id)
         raise VoxelizationError(f"voxelization of {id} failed")
     nchannels = prot_vox.shape[1]
     prot_vox_t = (